Related Art
A. CDK-Cyclin Complexes
Cell division is a complex process that is regulated by a number of cellular and environmental factors. Recent studies have identified two classes of proteins that appear to play key roles in controlling the cell cycle. These classes include the cyclin-dependent protein kinases ("cdks") and the cyclins. Cdks function by phosphorylating selected protein substrates in the cell; these phosphorylated proteins in turn "signal" the cell to either enter or continue the process of cell division. For cdks to be active, i.e., to phosphorylate other proteins, they must be bound to a cyclin protein. Thus, cyclins "regulate" the activity of cdks by binding to them.
Several cdks and cyclins have been identified in mammals. At present, 9 cdks are known, and they are referred to as "cdk1", "cdk2", and so on. Ten families cyclins are currently recognized, and are referred to as "cyclin A", "cyclin B", and so forth through "cyclin J". For a general review of cyclins see Coats et al. (in Signal Transduction, Heldin and Purton, eds., Chapman and Hall, publishers {1996}; pages 347-360) and Lees (Curr. Opinions Cell Biol., 7:773-780 [1995]).
Each cyclin family may have more than one member. Mammals, for example, have two types of cyclin A; cyclin Al and cyclin A2, and three types of cyclin D (D1, D2, and D3). Prior to the present invention, only one mammalian cyclin E, cyclin E1, was known, although Zariwala et al. (Pathways to Cancer, a Cold Spring Harbor Winter Conference, Harlow et al., eds., Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y. p. 49 [1998]) have purportedly identified a novel cyclin E protein. No DNA or amino acid sequence data regarding this molecule is available.
Cyclin family members are typically classified based on their amino acid sequence homology to existing family members. For example, cyclin A2 has about 45 percent amino acid sequence homology to cyclin Al, but only about 19 percent sequence homology to cyclin D1 and 21 percent sequence homology to cyclin E1 (as calculated using the MacVector.RTM. clustal alignment software program from Oxford Molecular Group).
All cyclin molecules contain an amino acid sequence domain referred to as the "cyclin box". The cyclin box is about 100 amino acids in length (the average full length of a cyclin polypeptide is 300-500 amino acids) and is located in the middle portion of each cyclin. While the precise amino acid sequence of the cyclin box varies from family to family, and even within members of a family, there is a highly conserved motif within the cyclin box that is consistently present in all cyclin boxes (see Prosite public database, accession number PS00292 which sets forth a cyclin box "consensus" sequence).
Cyclins and cdks bind to each other in a highly selective manner; not all cyclins bind to all cdks. For example, cyclin D1 can associate with cdk4, but not with cdk2. Similarly, cyclin E can associate with cdk2 and cdk3, but not with cdk4; cyclin A can associate with cdk2, but not with cdk5. Formation of cyclin-cdk complexes is transient; the two molecules may be present in the cell at the same time, but can only form an active complex if cdk is phosphorylated by an enzyme referred to as "cak" for cdk activating kinase.
B. Cyclin E1
Human cyclin E1 was first cloned in 1991 and was found to bind to and activate cdk2 (U.S. Pat. No. 5,449,755 issued Sep. 12, 1995; WO 93/06123 published Apr. 1, 1993; Koff et al., Cell, 66:1217-1228 [1991]; see also PCT patent application WO 98/03649, published Jan. 29, 1998). Cyclin E1 homologs have been identified in Drosophila (Richardson et al., Development, 119:673-690 [1993]), mouse (Damjanov et al., Biochem. Biophys. Res. Comm., 201:994-1000 [1994), Xenopus (Chevalier et al., J. Cell Sci., 109:1173-1184 [1996]) and Zebrafish (Yarden et al., Devel. Dynam. 206:1-11 [1996]).In addition, two cyclin E1 variants have been reported. The first of these is a human cyclin E1 splice variant (Mumberg et al., Nuc. Acids Res., 25:2098-2105 1997]) that purportedly has an internal deletion of 45 amino acids, and has an expression pattern that is distinct from full length cyclin E1. The other reported variant purportedly lacks the 15 amino acids at the amino terminus (Ohtsubo et al., Mol. Cell. Biol., 15:2612-2624 [1995]).
A novel cyclin polypeptide, called cyclin N, that purportedly is related to cyclin E1 has recently been reported (Lauper et al., Abstracts from the 1998 Cold Spring Harbor Laboratory Cell Cycle Meeting, Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y. [1998] p. 115).
C. Cyclins and Cancer
One hallmark of cancer is the uncontrolled, and seemingly unregulated, division of cells. The role of cyclins and cdks in regulation of cell division suggests that these proteins may be involved in converting normal cells to cancerous cells. Numerous recent studies have thus focused on cyclin and cdk involvement in cancer. For example, in a recent review article, Sherr (Science, 274:1672-1677 [1996]) pointed out that overexpression of cyclin D1 is seen in sarcomas, colorectal tumors, and melanomas, and that cyclin E is overexpressed in breast, stomach, colon and endometrial carcinomas.
Sarcevic et al. (J. Biol. Chem., 272:33327-33337 [1997]) describe the substrate specificies of various cyclin-cdk complexes in T-47D human breast cancer cells. They found, for example, that cyclin D1-cdk4 phosphorylated a 38 kDa protein, while cyclin D3-cdk4 phosphorylated a 105 kDa protein, a 102 kDa protein, and a 42 kDa protein. Cyclin E1-cdk2 and cyclin A-cdk2 phosphorylated several proteins.
Cyclin E1 has been implicated in breast cancer. In young breast cancer patients, high cyclin E1 expression in breast tumor tissue purportedly correlates with decreased survival (Porter et al., Nature Med., 3:222-225 [1997]). In addition, cyclin E1 is apparently overexpressed in various breast cancer cell lines (Gray-Bablin et al., Proc. Natl. Acad. Sci USA, 93:15215-15220 [1996]).
Given that cyclin-cdk complexes are involved in cell division and are active in cancer cells, inactivation of these complexes could result in decreased tumor cell proliferation. U.S. Pat. No. 5,645,999 issued Jul. 8, 1997 describes assays that are purportedly useful for identifying compounds that modulate cyclin E1 activity.
A few naturally occurring proteins have been identified as inhibitors of cyclin-cdk activity. These include the proteins p21 and p27 (U.S. Pat. No. 5,688,665 issued Nov. 18, 1997; PCT WO96/02140, published Feb. 1, 1996; Sherr et al., Genes and Devel., 9:1149-1163 [1995]). The protein p21 purportedly binds directly to cdk1, cdk2, and cdk4, and can apparently inhibit both cyclin D-cdk complexes and cyclin E-cdk complexes (Sherr et al., supra). The protein p27 purportedly binds to cyclin-cdk complexes (rather than to isolated cdks) and can apparently inhibit cyclin A, B, D and E dependent kinase activity (Sherr et al., supra). However, cyclin E1-cdk2 has also been found to purportedly regulate p27 (Sheaff et al., Genes and Devel., 11:1464-1478 [1997]).
In view of the devastating effects of cancer, there is a need in the art to identify molecules in the human body which may have an important role in the etiology of this disease, and to manipulate the expression of such molecules in patients suffering from these and related diseases.
Accordingly, it is an object of this invention to provide nucleic acid molecules and polypeptides that have a role in cell division.
It is a further object to provide methods of altering the level of expression and/or activity of such polypeptides in the human body. related objects will readily be apparent from a reading of this disclosure.